1) The document discusses recent developments in technologies for processing molten aluminum. It outlines the key steps in molten metal processing including crucible pre-treatment, furnace processing, and in-line treatment like degassing and filtration. 2) Recent progress has been made in improving furnace processing technologies like rotary gas injection that increase contact between the metal and fluxing agents, improving impurity removal. However, furnace processing alone cannot achieve the cleanliness required for high-quality products. 3) Ongoing challenges remain in developing simple, accurate, and low-cost on-line tools for assessing metal cleanliness and controlling processing technologies to optimize quality, productivity, costs and environmental impact.

1. A TECHNICAL PERSPECTIVE ON MOLTEN ALUMINUM PROCESSING
Peter Waite
Alcan International Limited
Arvida Research and Development Center
Jonquière, Québec, Canada G7S 4K8
Abstract However, notwithstanding recent advances in metal quality, the
cast house metallurgist still faces significant challenges. Metal
In today's context of global competitiveness, all factors related to processing and monitoring costs are significant – available
molten metal treatment which directly or indirectly affect product resources must be used judiciously. Pressure to increase
quality, the environment and processing costs must be optimized. productivity and metal throughput are incessant, while at the same
In this regard, technology and innovation play a decisive role for time, metal quality variation is unacceptable. Environment issues
the development and implementation of the most appropriate must be taken seriously and can limit the metal processing options
molten metal treatment processes and practices. The following that are available.
discussion will review the most recent significant developments in
the field of molten aluminum processing and outline potential Recent progress made in the field of molten metal treatment
areas for improvement. technologies is summarized and future challenges are suggested.
Introduction Overview of Metal Processing Steps
In order to meet the continually increasing product performance Many different metal treatment technologies and practices are
requirements of the world market place, molten metal quality is a available to the cast house metallurgist. The successful
major preoccupation of the cast house. In the context of this implementation and use of these technologies is accomplished
discussion, metal quality refers to the degree to which an only when an appropriate balance is achieved between metal
aluminum alloy is free of the following contaminants: alkalis quality, productivity, cost and the environment. The general
(sodium, calcium, and lithium), non-metallic inclusions, and sequence of molten metal processing steps is shown in Figure 1
dissolved hydrogen. Improvements in metal quality have been and consists of crucible pre-treatment upstream of the furnace,
achieved with the development and implementation of molten furnace processing, and in-line treatment – degassing and
metal treatment technologies during the last 20 years. Advances, filtration. Additionally, the transfer of the liquid metal between
particularly during the last decade, are significant and reflect an each processing step must be given particular attention. It is
improved comprehension of the underlying principles governing generally recognized that as the liquid metal advances towards the
molten metal treatment. ingot, metal treatment operations become increasingly "critical".

2. Electrolytic Cell
Crucible
Pre-treatment
Casting
Casting Furnace Degasser Filter
Melting Furnace
Figure 1: Sequence of Molten Metal Processing Steps.
In the past, process development and optimization efforts focused Motivated by cost reduction or simply by a sometimes limited
on individual metal treatment steps. Semi-empirical methods, availability of good quality clean scrap metal, today's aluminum
requiring little knowledge of the underlying metal treatment recycling plants must have increased flexibility to accept variable
mechanisms, were used to relate process changes to final product scrap quality, and as such, the molten metal treatment system must
performance. This situation was complicated by the low impurity be increasingly robust to handle a higher and fluctuating impurity
concentrations that had to be accurately measured and because the load. Similarly, present trends in aluminum smelter operation
available metal quality measurement techniques, at the time, were include efforts to improve current efficiency and to reduce
costly and often laboratory-based - leading to limited acceptance environmental emissions. This has led to increasingly high
or use by the industry. sodium and/or lithium levels in the primary metal produced.
Consequently, improved molten metal treatment performance in
the cast house is necessary.
Molten Metal Analysis and Control
Control Technologies
Impurities
Quantitative instrumental techniques capable of accurately
The types and sources of impurities present in liquid aluminum, as measuring the extremely low concentrations of the various
well as their detrimental effects on specific products, have been contaminants are presently used throughout the aluminum
reviewed in detail elsewhere(1). It is pertinent to note that the industry.
molten aluminum supplied to the cast house comes from two
distinctly different sources: smelter electrolysis and remelt/recycle Inclusions – Metal Cleanliness Considered industry standards, the
operations. The impurities present in the metal supplied from LiMCA and PoDFA metal cleanliness assessment techniques,
these sources are also different and can affect the metal treatment developed by Alcan, have been previously described in
strategy that is used. Remelted metal is normally associated with detail(2, 3, 4). Briefly, LiMCA is based on the resistive pulse
higher levels of hydrogen, calcium, and hard oxide inclusions that principle and generates both an inclusion concentration value and
are formed during high temperature scrap melting processes. On a complete size distribution. PoDFA is based on the filtration of
the other hand, smelter metal is associated with higher levels of 1-2 kg of metal through a well-calibrated small filtration disk with
sodium, aluminum carbide inclusions, as well as non-metallic subsequent metallographic examination of the inclusion
inclusions generated from the addition of large quantities of concentrate. Information on inclusion species and sizes, as well as
alloying elements. Table I summarizes the impurity levels present a semi-quantitative total inclusion level expressed in mm2/kg are
in the metal supplied to the cast house. obtained. LiMCA and PoDFA are complementary technologies in
as much as their respective quantitative and qualitative capabilities
Table I Typical Impurity Levels in Metal from Smelter and combine to give essential data required for an informed judgment
Remelt Sources of metal cleanliness.
Characteristic Smelter Remelt However, significant challenges remain to be overcome. The
Alloyed or close to LiMCA instrument is on-line, but complex to operate, and PoDFA
Composition ≥ 99.7% Al analysis requires metallographic expertise/facilities and is thus
final composition
off-line. An inherent weakness of the LiMCA technique is the
Hydrogen 0.1 – 0.3 ppm 0.2 – 0.6 ppm inability to distinguish the physical state of the inclusion.
Consequently, and unless a deep bed filter is used, reliable
Alkali Na 30 – 150 ppm ≤ 10 ppm LiMCA measurements are difficult to obtain immediately
Ca 2 – 5 ppm 5 – 40 ppm downstream of an in-line degasser(5) due to the complex
Li 0 – 20 ppm < 1 ppm interactions between the gas/liquid/solid inclusions present.
Exploratory development work on an alternative inclusion
0.5 < mm2/kg < 5.0
Inclusions > 1 mm2/kg detection technology focuses on the use of ultra-sound(6). It
Al2O3, MgO,
(PoDFA scale) Al4C3 remains that a simple, low cost yet quantitatively accurate on-line
MgAl2O4, Al4C3, TiB2
metal cleanliness assessment tool is needed by the industry.

3. Hydrogen The AlSCAN™ technology(7, 8) is now in use through- Molten Metal Processing Technologies
out the industry and is recognized as the standard technique for
on-line hydrogen analysis. Due to it's high precision, AlSCAN Over the last 10 years, the need to produce higher quality products
has made possible the quantification of different phenomena (such has encouraged nearly uninterrupted development efforts in the
as ambient humidity) affecting hydrogen in molten aluminum. field of metal processing technologies. The clear benefits of these
technologies with respect to quality, cost, productivity and the
Chemical Composition Optical emission spectroscopy has environment, has led to their widespread use throughout the
established itself as the reference analysis technique for the aluminum industry.
measurement of alloy chemical composition at the level of major
constituents, as well as trace elements. A complex calibration and To appreciate the overall molten metal processing performance
alloy matrix correction procedure is required for the different alloy that is now demanded in the cast house, Figure 3 summarizes the
families, and remains a technical obstacle to overcome. low impurity levels that must be achieved for various products.
Compared to the incoming impurity levels (Table I), highly
Control Strategies efficient impurity removal is required during metal processing.
Measurement techniques capable of quantifying all aspects of
metal quality have been in widespread use for roughly 10 years.
3
Alkalis
With the specific objective of process development, focused Inclusions Parts per
campaigns in the cast house involving metal quality 1 million
measurements, before and after the different stages of metal 1000
Unfiltered
treatment, have permitted the identification of important process
Hydrogen
parameters. Combined with heightened efforts in metallurgical
Alloy
process modeling, an improved understanding of the underlying 100
Extrusion
Pure
metallurgical principles has been achieved, allowing advances in Parts per
the development and optimization of metal treatment technologies billion
and practices. 10
Filtered
The considerable measurement costs related to these process
Foil
Filtered
development activities were easily justified by the 1
Alloy
Computer
performance/quality improvements that were progressively 1000
obtained. However, at present, molten metal treatment
Disks
technologies have reached a state of maturity that no longer
100 Parts per
requires such a high level of quality measurement support. trillion
The most effective use of metal quality measurement resources is
being re-examined. A strategy focusing on product quality control 10
and process capability monitoring is presently emerging.
Statistical process control techniques are being exploited to Figure 3: Typical Impurity Concentrations in Some Aluminum
maximize the amount of information extracted from the data. The Products.
use of control charts and the establishment of process control
limits enable the metallurgist to learn from the process in order to Crucible Pre-treatment
improve it. Finally, process benchmarking and data exchange
between different divisions of a company facilitates the Several crucible-based treatment technologies(9, 10, 11) have been
establishment of process "best practices". Figure 2 summarizes developed with the objective of removing alkalis and inclusions
this metal analysis and control strategy. from the metal prior to transfer into the casting furnace. All these
technologies are similar in as much as a reactive flux (either
chlorine-based or salt-based) is injected into a well-stirred crucible
MEASURE of liquid aluminum. The present consensus favors the chlorine-
free technology for both performance and environmental reasons.
CONTROL / COMPARE Due to the crucible shape (surface/volume ratio), the metal depth
and the intensity with which the metal is stirred, crucible pre-
treatment is rapid, environmentally sound and significantly
reduces the need for subsequent furnace fluxing.
UNDERSTAND
In the past, crucible pre-treatment technology was applied almost
exclusively to smelter applications. More recently, and as the
IMPROVE performance of subsequent furnace treatment technologies have
improved, the need for crucible pre-treatment has been re-
Figure 2: Metal Quality Measurement and Control Strategy. questioned with respect to its costs versus its advantages.
However, in the present context of increasing alkali impurity
levels in primary smelter metal, and the need to accept potentially
lower quality scrap in recycling operations, the beneficial impact

4. of crucible pre-treatment can only become more significant for cleanliness, alkali removal, as well as dross formation, have been
both smelter and recycle-based cast houses. Clearly, the challenge quantified for a wide variety of alloys. In the past, limited
is to readily identify situations where the benefits of crucible pre- improvements to lance fluxing were achieved simply by the use of
treatment can be best exploited. With more widespread use more lances or by the introduction of porous plugs (with the
envisioned, molten metal management may also become an issue. associated operational difficulties).
Finally, extension of this concept to include alloy preparation has
the proven potential(12) for further increasing the value of this Recent improvement to furnace treatment effectiveness has been
metal processing step. achieved by "process intensification" aimed at increasing metal
stirring and fluxing kinetics. Metallurgical modeling has
Furnace Treatment confirmed(19) that the rate limiting step of the furnace fluxing
process, assuming the presence of a minimum of bulk metal
The most common type of casting furnace in current use circulation, is the gas/liquid interfacial contact area generated by
throughout the aluminum industry is the fossil fuel heated the gas purging system.
reverberatory design. The reverberatory furnace, containing a
relatively shallow melt depth, with a high surface area to volume The rotary gas injection (RGI) furnace fluxing technology,(20, 21, 22)
ratio, is optimized for heat transfer. This melt geometry, utilizing high shear gas dispersers, has been know for some time.
combined with the lack of bulk metal circulation, does not This furnace fluxing technology, and others(23), markedly
promote efficient metallurgical treatment. In addition, the vertical increases the interfacial contact area between the liquid metal and
temperature gradient that is generally established through the melt the fluxing agent. Recent adaptation of this technology by the
can be greater than 200oC and is responsible for elevated dross aluminum industry has resulted in significant gains in the cast
formation and reduced heat transfer. These problems are house. Increased impurity removal rates have reduced furnace
magnified as furnace sizes constantly increase. cycle times and improved productivity. The absolute metal
quality levels achieved have improved, lessening the load on
Conventional furnace fluxing with static lances is generally subsequent in-line metal processing steps. Equally as important,
accepted as a rather inefficient way to treat molten aluminum, and significant reductions in chlorine consumption, made possible by
depending on the metal quality requirement, excessively long the improved process efficiency, have had a direct and positive
treatment periods are necessary. This has a negative impact on impact on atmospheric emissions.
both productivity and the environment.
Salt fluxes can be used effectively for the replacement of chlorine
In an effort to overcome reverberatory furnace design limitations, during furnace fluxing(24). Rotary flux injection technologies have
considerable effort has been invested by the aluminum industry to also provided the opportunity to replace chlorine with a salt-based
understand the fundamental metallurgical phenomena taking flux(25, 26) while reducing environmental emissions and
place(13, 14, 15, 16) and to apply this understanding to improve maintaining metallurgical treatment performances comparable to
processing technology and practices. During the last decade, two chlorine. Figure 4 shows the impact of these advances on the
methodologies have been pursued to overcome the inherent overall chlorine consumption of a cast house.
inefficiencies of furnace treatment: first, by directly improving
the efficiency of furnace treatment itself, and secondly, by
reducing the need for furnace treatment and exploiting the more 0.3 - 0.7
Kg/t
efficient crucible pre-treatment and/or in-line treatment
technologies. 0.1 - 0.2
Kg/t
Stirring In the past, stirring by lance gas injection was
0.05
commonplace and suffered from limited effectiveness as well as Kg/t
increased dross generation. Today, advanced subsurface stirring 0
Kg/t
technologies include: pneumatic jet stirring, electromagnetic
approaches and rotary devices.
The metallurgical and operational advantages of forced metal Lance
circulation have been previously reported in detail(17) and include Furnace (Chlorine) RGI RFI
Fluxing (Chlorine) (Salt)
rapid homogenization of the alloy, reduction of the temperature Target
Ò0Ó
gradient and the associated metal oxidation, increased uniformity In-Line Chlorine Chlorine
Chlorine
and consistency of metal processing and improvements to furnace Fluxing Chlorine
control in general. The use of bulk metal circulation systems has
grown throughout the industry.
Figure 4: Progressive Reduction of Chlorine Consumption in the
Fluxing Furnace "fluxing" refers to the removal and separation of Cast House.
impurities from liquid metal by direct reaction and dewetting
based on the use of a chemically active agent. Until recently, the However, there remain important challenges to overcome with
aluminum industry relied exclusively on the use of chlorine for respect to furnace treatment. The casting furnace is not an
fluxing. efficient chemical reactor. Although present design limitations
have been partially overcome by the introduction of improved
The complex mechanisms of furnace fluxing with static lances and fluxing technologies, furnace fluxing cannot provide metal that
chlorine have been studied in detail(18) and its effects on metal meets the cleanliness requirements of critical products.

5. Subsequent in-line metal processing is still necessary. There is a the unit between casts and the resulting metal loss at alloy change.
general need to reduce metal surface turbulence in the furnace that The large floor area requirement and equipment complexity is also
leads to dross formation and metal loss. The majority of dross is becoming a very important issue for the cast shop.
generated during metal transfer into the furnace, with the
remaining dross generated during other furnace processing To address these issues, a recent in-line degassing innovation is
activities, including alloy addition and skimming. Employing the the development of a trough-based process(31, 32, 33) targeting multi-
best technologies available today, including siphon metal transfer, alloy cast houses. The trough-based process virtually eliminates
a target dross generation rate of approximately 1% is possible. metal loss at alloy change and maintains roughly equivalent metal
However, siphoning cannot be used in all cast houses, thus treatment performances and capacities. Development of the Alcan
underlining the need for continued efforts in the area of furnace Compact Degasser technology in particular, was based on the use
design for improved metal transfer. Finally, the trend towards of a hydrogen diffusion model(34) and is a prime example of how a
casting furnaces with greatly increased capacities severely limits fundamental understanding of the metallurgical principles can lead
production flexibility with respect to batch size and alloy change. to process innovation.
In-line Metal Treatment The use of high shear rotary injectors with chlorine-containing gas
mixtures for the treatment of Al-Mg alloys leads to the formation
Due to the inherent inefficiencies of furnace treatment, a growing of minute magnesium chloride inclusions. If an effective metal
emphasis has been placed on the utilization of in-line metal filtration system is not used, the entrainment of these inclusions
processing over the last 20 years. Many in-line treatment downstream of the degassing unit tends to accelerate surface
processes have been developed with the objective of providing oxidation of the molten metal. Oxide-related production or ingot
more efficient and consistent metal treatment performance, while quality problems can result. This phenomenon can limit the
reducing treatment time, emissions and metal loss due to dross amount of chlorine that can be used during in-line degassing, the
formation. There are two categories of in-line metal treatment consequence of which is poor alkali removal performance. Work
processes namely, degassing/fluxing and filtration. has been initiated on separation techniques for the removal of
chloride inclusions(35, 36) in an effort to overcome this limitation.
Degassing The in-line degassing process is now widely accepted
and used almost universally throughout the aluminum industry. The elimination of chlorine from the cast house is advantageous
During the degassing treatment, a gas composed of argon and for safety and environmental reasons, as well as for reduction of
chlorine is dispersed into the liquid metal using one or more high- maintenance costs. Proven alternatives to the use of chlorine exist
speed rotary injectors. The in-line degassing process design is for both crucible pre-treatment and furnace fluxing. Using these
based on the classic multi-stage stirred tank reactor and on salt-based alternatives, the overall chlorine consumption in the
recognition of the need to generate small gas bubbles to increase cast house can be drastically reduced without penalizing metal
the interfacial gas-metal contact area. Due to their functional quality or productivity. The last technical barrier to a truly
similarity, all rotary-type in-line fluxing processes provide chlorine-free cast house is the development of a chlorine
roughly equivalent metal treatment performance. It should be alternative for in-line degassing. However, no such chlorine
noted that although these processes are called "degassers", they alternative presently exists. Today, the only option is to operate
function to remove not only hydrogen but inclusions and alkali the degasser without chlorine while accepting reduced
impurities as well. The physical and chemical factors which metallurgical performance - albeit at the same time, solving the
influence in-line degassing/fluxing performance have been magnesium chloride entrainment problem. In this context, and
published in detail elsewhere.(27, 28) keeping in mind the integrated approach to metal treatment, zero
chlorine operation for in-line degassing is a potentially viable
Recent advances in the design of in-line degassers have resulted in solution for achieving a chlorine-free cast house. Upstream and
more efficient and consistent metallurgical performance. downstream processing practices/technologies would have to be
Improved rotary gas dispersers with increased shearing(29) altered to maintain the overall metal treatment performance of the
maximize gas-metal contact area – a necessary condition for system.
efficient metal treatment.
Tighter environmental emission limits are issues that must be
The capacity of in-line degassing technologies has been scaled up addressed. Will technological improvements or operational
by the use of multiple staging, in order to meet the demands of changes be sufficient to meet the environmental regulations of the
increased casting rates. Multiple staging also has the advantage of future? Perhaps a more drastic re-thinking of the in-line degassing
having "built-in redundancy". The impact of a rotor failure on the process will be required.
metal quality is less severe. Another recent improvement is the
operation of in-line degassers in a "sealed" mode,(30) that is to say, Filtration In the aluminum industry, molten metal filtration has
with an inert cover gas – essentially free of oxygen – over the been used in one form or another for many years(37). It is common
liquid metal surface inside the unit. The impact of this to use a filter strictly for precautionary reasons, that is to say, as a
development on the removal of impurities has not been fully "safety net" to guard against uncontrolled upstream process
quantified. However, there are definite advantages in terms of variations that result in sporadically high inclusion levels.
considerably reduced dross generation, easier operation However, in the context of this discussion, the use of a filter is
(skimming less frequently) and possibly lower particulate considered an essential metal processing step. The objective of
emission rates. filtration is to improve the metal quality to the degree required for
demanding products/applications.
On the other hand, scale-up of the conventional multiple-stage
degasser has greatly increased the amount of metal retained inside

6. Unfortunately, a poor scientific understanding of the fundamental necessary to filter hundreds or thousands of tons of metal between
mechanisms taking place during aluminum filtration has hampered media changeover. The large flow area of these filters reduces the
developments in filter design. Empirical design methods are often metal velocity and greatly improves the filtration efficiency at the
used based on trial and error. The filter was considered a "black expense of a high metal holdup volume and a high floor space
box" because it was almost impossible to obtain a direct measure requirement. The elevated cost associated with filter media
of what the filter was doing. At best, physical model analogues changeover limits the use of multi-cast filters to cast houses
(water models) were used as a design guide. having infrequent alloy changes.
Two technical advances have started to change this situation. Regardless of whether a deep bed filter, a ceramic foam filter
First, is the development of inclusion measurement capabilities. (CFF), or a cartridge filter is used, the technical challenge facing
For example, the use of LiMCA to quantitatively measure metal metal filtration has always been defining the optimum balance
cleanliness has resulted in a practical comprehension of filter between the required metal cleanliness, the filtration cost, the
behavior in terms of operational phenomena such as inclusion available head loss and more recently with significantly increased
release events, maximum filter loading and the impact of metal casting rates, the size of the filter.
velocity on filtration efficiency. Secondly, and more recently, is
the scientific advancement in fundamental research that has There is a definite need to develop an efficient, low hold-up
started to elucidate the important mechanisms that are active filtration process capable of treating high metal flow rates.
during aluminum filtration.(38, 39, 40) However, continued effort is
required to gain a more complete understanding of depth filtration, Metal Conveying
the knowledge of which can eventually serve as the basis for more
sophisticated filter design techniques. Particular attention must be given to the metal transfer equipment
that connects the elements of an integrated metal treatment
What is understood today? Filtering techniques employed in the system. Metal turbulence and cascading must be avoided. The
aluminum industry are mostly based on the depth filtration mode resulting oxides that are formed can undo much of the metal
which involves the deposition and retention of inclusions cleanliness gains that were achieved during upstream processing.
throughout the entire depth of the filter. Inclusions are removed In addition, it is recognized that the refractory and patching
from the metal stream by sedimentation and are very weakly materials used in troughs (and in-line treatment units) can lead to
retained on the surface of the filter medium. Filtration the formation of heterogeneous inclusions. If this occurs late in
performance is improved by: the sequence of metal processing, the effects could be highly
detrimental on product quality. Even ambient humidity can cause
providing a stable and wetted interfacial contact between problems and has been shown to cause hydrogen pick-up in liquid
the liquid metal and the filter medium; metal exposed to air. Under conditions of high ambient humidity,
reducing pore size; products with strict hydrogen limits can only be cast with great
increasing filter depth; difficulty.
reducing the metal flow velocity.
The basic trough design has not changed appreciably for decades.
A major constraint of aluminum filter design is the operational Unlike the magnesium industry, where elimination of the metal
limitation of the very low head loss that can be accepted across the free-surface has been accomplished using closed liquid metal
filter(41). Note that the factors that improve filtration performance conveying systems, the reactive/aggressive nature of liquid
(above) all tend to either increase the head loss or increase the aluminum has prevented equivalent developments in this industry.
filter size – both of which are undesirable. Finally, more work is Upon completion of metal processing, there is a need to maintain
needed to relate molten metal cleanliness to product properties and the liquid metal at its highest quality level for delivery to the
performance. In many cases, the real metal cleanliness level that casting machine. The present trough design cannot always fulfil
is needed for a particular product is not properly quantified. this requirement.
Note that cake mode filtration is not presently used in the Future Implications
aluminum industry. Cake filtration involves the deposition of a
layer of inclusions at the inlet to the filter medium with little or no The integrated approach to metal treatment that has just been
penetration of the inclusions into the internal pore structure of the described has gone through a period of process development and
filter. This results in a very rapid metallostatic head build-up and optimization. Each element of the treatment sequence is being
is unacceptable for cost and practical operating reasons. pushed to its maximum. Although there remain challenges, only
step-wise improvements to quality/productivity/capacity are
Single-use filters such as the ceramic foam filter (CFF) are anticipated.
physically small and are changed after every cast. This gives the
advantage of low floor space and low metal holdup - facilitating In order to make a "quantum step" change, future technical
alloy changes. However, because of their relatively small hurdles may best be overcome by a complete re-thinking of the
dimensions and the high metal velocity through the filter, molten metal treatment strategy in the cast house. It must be
metallostatic head loss considerations limit the thickness and the recognized that multiple interventions are presently required to
minimum practical filter pore size that can be used. As a remove some impurities. For example, it is necessary to use all
consequence, these filters are not highly efficient(42). the metal processing steps to remove non-metallic inclusions,
while alkali removal is done during two or three steps. The future
On the other hand, deep bed filters are used during multiple casts. challenge may be to re-think metal treatment technologies,
They are physically large to contain the amount of filter media including the casting furnace, starting from a "carte blanche" - the

7. premise being that a metal processing step should be done once, 3) J.-P. Martin, R. Hachey, and F. Painchaud, "On-line Metal
but it must be done "right". For example, if inclusion-free metal Cleanliness Determination in Molten Aluminum Alloys Using
could be produced in the casting furnace, in-line treatment for the LiMCA II Analyser" Light Metals, 1994, 915-920.
inclusion removal would no longer be necessary. Conversely, if a
"super" filter could be developed capable of accepting very high 4) D. Doutre et al., "Aluminium Cleanliness Monitoring:
inclusion loading, upstream metal processing for inclusion Methods and Applications in Process Development and
removal would no longer be necessary and significant productivity Quality Control", Light Metals 1985, TMS-AIME, 1179-1195.
gains and/or processing cost reductions could be achieved. 5) H. P. Krug, and W. Schneider, "A Contribution to Inclusion
Measurement After In-line Degassers with PoDFA and
LiMCA", Light Metals, 1998, 863-870.
Conclusions
6) I. D. Sommerville, N. D. G. Mountford, and L. C. B. Martins,
There still remain key molten metal treatment challenges in the "Laboratory and Industrial Validation of an Ultrasonic Sensor
cast house. for Cleanliness Measurement in Liquid Metals", Light Metal,
2000, 721-726.
Molten metal processing must be more robust – capable of 7) J.-P. Martin, F. Tremblay, and G. Dubé, "Alscan: A New and
accepting poorer quality metal at the input, while producing Simple Technique for In-line Analysis of Hydrogen in
higher and less variation metal quality at the output. Aluminum Alloys", Light Metals, 1989, 903-912.
There is a need to develop a more selective on-line inclusion 8) C. Dupuis et al., "An Analysis of Factors Affecting the
detector. Response of Hydrogen Determination Techniques for
Aluminum Alloys", Light Metals, 1992, 1055-1067.
Further development is needed to optimize furnace design
with respect to processing performance and batch size 9) G. Dubé, and V. J. Newberry, "TAC – A Novel Process for the
flexibility. Removal of Lithium and Other Alkalis in Primary
Aluminum", Light Metals, 1983, 991-1003.
The last technical barrier to a truly chlorine-free cast house is
the development of a chlorine alternative for in-line 10) B. Rasch, E. Myrbostad, and K. Hafsas, "Refining of Potroom
degassing. Metal Using the Hydro RAM Crucible Fluxing Process", Light
Metals, 1998, 851-854.
Continued effort is required to gain a more complete
understanding of depth filtration mechanisms. 11) F. Achard, and C. Leroy, "Pre-treatment in Potlines Crucibles:
The Mixal Process", Light Metals, 1990, 765-768.
More work is needed to relate metal cleanliness requirements
to product properties and performance. 12) B. Gariépy et al., "Aluminum Ladle Metallurgy (ALM): A
New Process for More Efficient Alloy Preparation", Light
The development of an efficient, low holdup filtration Metals, 1988, 457-462.
process capable of treating high metal flow rates is needed.
13) B. Kulunk and R. Guthrie, "On the Kinetics of Removal of
Improvements are needed for liquid metal conveying systems Sodium from Aluminum and Aluminum-Magnesium Alloys",
to prevent dross formation and metal loss. Light Metals, 1992, 963-975.
14) G. K. Sigworth, "Gas Fluxing of Molten Aluminum", Light
Acknowledgements Metals, 2000, 773-778.
15) R. R. Roy, T. A. Utigard, and C. Dupuis, "Inclusion Removal
The author wishes to thank Alcan International Limited for the During Chlorine Fluxing of Aluminum Alloys", Light Metals,
permission to publish this article. Discussions and suggestions 1998, 871-875.
during the preparation of this article, made by Messrs. Ghyslain
Dubé, Claude Dupuis and Dr. Jean-Pierre Martin, are also 16) R. R. Roy, T. A. Utigard, and C. Dupuis, "Inclusion Removal
gratefully acknowledged. Kinetics During Chlorine Fluxing of Molten Aluminum",
Light Metals, 2001, 991-997.
17) M. A. Thibault, J. C. Pomerleau, and F. Tremblay, "Molten
References
Metal Stirring: The Alcan Jet Stirrer", Light Metals, 1991,
1005-1011.
1) "Principles of Aluminum Metal Treatment", Continuing
18) C. Celik and D. Doutre, "Theoretical and Experimental
Education Committee Short Course, Presented at the 122nd
Investigation of Furnace Chlorine Fluxing", Light Metals,
TMS Annual Meeting, Denver, Colorado, 21-22 February
1989, 793-800.
1993.
19) J.-F. Bilodeau, C. Lakroni, and Y. Kacaefe, "Modeling of
2) R. I. Guthrie, and D. A. Doutre, "On-Line Measurements of
Rotary Injection Process for Molten Aluminum Processing",
Inclusions in Liquid Metals", International Seminar on
Light Metals, 2001, 1009-1015.
Refining and Alloying of Liquid Aluminum and Ferro-Alloys,
Trondheim, Norway, August 1985, 145-163. 20) G. Béland, C. Dupuis and J.-P. Martin, "Improving Fluxing of
Aluminum Alloys", Light Metals, 1995, 1189-1195.

8. 21) L.I. Kiss, and J.-F. Bilodeau, "Flow Structure Around an 33) C. English, "The Cast House Trough Reactor A New Concept
Axial-flow Rotor for Gas Injection", Light Metals Processing for Degassing Reactor Design", Light Metals, 1998, 811-815.
and Applications, Canadian Institute of Mining, Metallurgy
and Petroleum, 2001, 1009-1016. 34) P. D. Waite, "Use of a Hydrogen Diffusion Model for
Prediction of Aluminum In-line Degasser Performance",
22) F. Boeuf, O. Baud, and C. Leroy, "Rotating Injectors in Recent Developments in Light Metals, Proceedings of the
Casthouse Furnaces", Light Metals, 1990, 755-768. International Symposium on Recent Developments in Light
Metals, The Canadian Institute of Mining, Metallurgy, and
23) R. S. Henderson, D. V. Neff, and C. T. Vild, "Recent Petroleum, Toronto, Ontario, August 20-25 1994, 431-440.
Advancements in Gas Injection Technology Using Molten
Metal Pumps", Light Metals, 2001, 1033-1039. 35) L. S. Aubrey et al., "New Product Developments for
Aluminum Cast Houses", Aluminium Cast House Technology,
24) D. DeYoung, "Salt Fluxes for Alkali and Alkaline Earth Proceedings of the 7th Australian Asian Pacific Conference,
Element Removal from Molten Aluminum", Aluminium Cast 2001, 23-43.
House Technology, Proceedings of the 7th Australian Asian
Pacific Conference, 2001, 99-113. 36) US Patent 5122184, "Molten Salt Coalescence in Molten
Aluminum", 1992.
25) G. Béland et al., "Rotary Flux Injection: Chlorine-Free
Technique for Furnace Preparation", Light Metals, 1998, 843- 37) R. Figueres, International Seminar on Refining and Alloying
847. of Liquid Aluminium and Ferro-Alloys, Trondheim, Norway,
August 1985, 323-332.
26) P. J. Flisakowski, et al., "Improvements in Cast Shop
Processing using Pyrotek's HD-2000 and PHD-50 Rotary 38) D. Kocaefe, R.T. Bui, and P. Waite, "2D Transient
Injector System", Light Metals, 2001, 1041-1047. Mathematical Model of Aluminum Filtration", Light Metals,
2000, 817-823.
27) T. A. Engh,., Principles of Metal Refining, Oxford Science
Publications, Oxford University Press, 1992. 39) D. Kocaefe, R. T. Bui and A. Chapdelaine, "1D Mathematical
Model of Deep Bed Filtration", Light Metals 1999, Canadian
28) G. K. Sigworth, and T. A.Engh, "Chemical and Kinetic Institute of Mining, Metallurgy and Petroleum, 1999, 203-217.
Factors Relating to Hydrogen Removal from Aluminum",
Metallurgical Transactions B, Vol. 13B, American Society for 40) D. Kocaefe et al., "Mathematical and Physical Modeling Tools
Metals and the Metallurgical Society of AIME, September for Aluminum Filtration", Light Metals 1999, Canadian
1982, 447-460. Institute of Mining, Metallurgy and Petroleum, 1999, 219-233.
29) J. E. Buehler, and R. A. Frank, "Startup and Evaluation of the 41) J. Szekely, Fluid Flow Phenomena in Metals Processing,
First SNIF SHEER System at Kaiser Aluminum, Trentwood Academic Press, New York, NY., 1979.
Works", Presented at the 122nd TMS Annual Meeting, Denver,
Colorado, Feb. 24, 1993. 42) C. Dupuis, G. Béland, and J.-P. Martin, "Filtration Efficiency
of Ceramic Foam Filters for Production of High Quality
30) J.-C. Terrier, "Alpur TS, Pechiney's New Global Concept For Molten Metal Alloys", Light Metals Processing and
Molten Aluminum Treatment", Light Metals, 1998, 779-783. Applications, Canadian Institute of Mining, Metallurgy and
Petroleum, 1993, 349-358.
31) P. Waite, and R. Thiffault, " The Alcan Compact Degasser: A
Trough-Based Aluminum Treatment Process Part 1:
Metallurgical Principles and Performance", Light Metals,
1996, 1001-1005.
32) P. D. Waite, "Improved Metallurgical Understanding of the
Alcan Compact Degasser After Two Years of Industrial
Implementation in Aluminum Casting Plants", Light Metals,
1998, 791-796.